Expose selected Crawlora endpoints as tool calls so agents can retrieve normalized JSON before summarizing or ranking results.
What you will build
Agent receives task
Planner selects Crawlora tool or endpoint
Crawlora returns normalized JSON
Agent summarizes, ranks, or compares
Store result or produce report
APIs used
Only endpoints that exist in the generated endpoint metadata are linked here. Missing optional endpoints are intentionally omitted.
POSTGoogleapiKey3 credits/request
Google search API
/google/search
Returns normalized Google web search results. Results are fetched through proxied browser renderers that race several concurrent renders per request and return the first clean result, with stale-cache fallback when available. The endpoint returns 503 when Google serves a challenge page or unusable HTML. Rate limit is enforced at 1 request per second, and if the limit is exceeded a 429 status code is returned with rate limit headers.
Searches JustWatch titles using the public credential-free website GraphQL endpoint. Country must be a two-letter ISO code such as `US`; language must be a two-letter code such as `en`.
Returns transcript segments for a YouTube video using YouTube's native player captions.
Set `format=text`, `format=srt`, or `format=vtt` to receive plain-text output instead of the standard response envelope.
Returns normalized product details for an Amazon ASIN on `amazon.com`, including pricing, availability, overview data, inline review samples, and descriptive content.
This is an illustrative agent tool schema, not an exact MCP schema. Use the MCP docs when a specific tool schema is available.
Illustrative tool schema
{
"name": "crawlora_google_search",
"description": "Search public Google results through Crawlora and return structured results.",
"parameters": {
"type": "object",
"properties": {
"query": { "type": "string" },
"country": { "type": "string" },
"language": { "type": "string" }
},
"required": ["query"]
}
}
Storage/output suggestion
Store tool inputs, response summaries, and source identifiers so agent outputs can be audited.
Error handling
Validate required inputs before calling Crawlora
Retry 429 and temporary 5xx responses with capped backoff
Log endpoint, input, timestamp, and request ID when present
Treat empty results as a state your application can handle
Open /docs/errors for production retry guidance
Rate-limit and credit planning
Estimate usage by multiplying requests by endpoint credit cost. The table below only shows real credit costs available from the billing constants.
Endpoint
Credit cost
Docs
Google search API
3 credits/request
/docs/google/google-search
Google Maps search API
5 credits/request
/docs/google-map/google-map-search
Search Nominatim places
1 credit/request
/docs/geocoding/geocoding-search
Search JustWatch titles
3 credits/request
/docs/just-watch/justwatch-search
Retrieve transcript for a YouTube video
3 credits/request
/docs/youtube/youtube-transcript
Search TikTok videos
8 credits/request
/docs/tiktok/tiktok-search
Search the App Store
1 credit/request
/docs/app-store/appstore-search
Search Google Play
1 credit/request
/docs/google-play/googleplay-search
Retrieve Amazon product details
2 credits/request
/docs/amazon/amazon-product
Search for products, users, or launches on Product Hunt
3 credits/request
/docs/product-hunt/producthunt-search
Production checklist
Keep API keys server-side
Use request timeouts
Back off on rate limits
Store raw responses or source IDs for auditability
Monitor credits and failures
Avoid unnecessary refreshes
Review responsible-use requirements
Responsible public web data workflows
Crawlora is designed for responsible structured public web data workflows. Customers are responsible for using Crawlora in compliance with applicable laws, third-party rights, target-platform rules, and Crawlora terms.